AI Transforms Auto Industry: Ford and Magna Lead the Charge
The automotive industry is increasingly using artificial intelligence to help choose which companies will supply components for vehicles, though some are wary of the potential impact on supplier relations.
The need to reduce the cost of building vehicles is critical as automakers fund investments for EVs and pursue profitability on those models. Companies are looking at opportunities to address those challenges, leading some to bet on AI as part of their toolkit.
Earlier this summer at the Aspen Ideas Festival in Colorado, Ford Motor Co. CEO Jim Farley suggested the automaker was looking at AI application in a number of areas: manufacturing to free up people to do "more valuable" work, diagnosing problems in vehicle software and improving safety-related systems.
"The biggest impact on the first inning of this is going to be our industrial supply chain," he said. "Identifying risk before humans can, helping us decide what is the best supplier for us. So much of the supply chain, which is super strategic at a company like Ford, a lot of that is just crunching numbers and making good decisions on data."
Ford didn't provide a subsequent interview after inquiries on the matter. Under the leadership of Liz Door, who became chief supply chain officer last summer, however, the automaker has reformed its supplier program to open the way for new partners and emphasize competitive quality, cost and delivery results.
Some other companies in the industry are already using AI in their supplier selection process. In recent months, Canada-based Magna International Inc., a provider of seating, electronics and vehicle integration services, has increasingly rolled out AI in its selection process.
It's still not regularly used in purchasing given Magna's longstanding familiarity with companies in the market and the practice of bundling multiple contracts to map out a vision with a few strategic partners, said Joerg Grotendorst, senior vice president of corporate research and development. AI's application, however, is helpful in addressing supply disruption and when the company is expanding into new product opportunities.
Joerg Grotendorst, Magna International Inc.'s senior vice president of corporate research and development, says AI can be valuable when a crisis arises.
AI can be helpful in identifying a replacement when, for example, a specialized aluminum manufacturing plant goes down and the company needs to supplement that component as quickly as possible.
"In the automotive industry, as we talk about homologation," Grotendorst said, "you have to find other suppliers with the same capability, same specification, which is a great task to be done by AI, because you have to fit some order concretely."
In other applications where there may be less familiarity with market players, AI can be helpful in taking an overwhelming amount of data and providing a broad overview that can be used to screen, compare and narrow the field to a few options for further discussion.
"Artificial (intelligence) for first reading can be helpful," Grotendorst said. "We found out, for example, that for (advanced driver assistance) systems and perception stacks — helping to sense the environment and identify objects in the environment — we have more than 1,000 companies found in China. Then to cluster them, then we use further tools."
The closer the process gets to production, however, the less useful AI becomes in its current state, he said: "What AI cannot find out really right now if, for example, we talk about new technologies, then you can look backward often if that company has been successful in launching new technologies, but that's not a grant to be successful in the future, as well.
"So, it's very important to sit together and have talks: What is the company's strategic target from our direction and from (the) supplier's direction? And then see if it matches and if then a new partnership can be established."
Development of a product can take two to three years, so any financial benefits of the rollout of this technology haven't yet been felt, Grotendorst said.
In pursuing business opportunities, suppliers don't see how automakers are using the data and information they supply, but given the increasing prevalence of AI, Grotendorst said he assumed if the customer is creating an overview of existing suppliers in a technology field that AI is going to be used more. That, though, hasn't necessarily changed how Magna is approaching new requests for proposals, especially given the strict standards and regulations that already have to be met in the automotive industry.
If it's increasingly up to a computer to pick out key points of data, it could result in suppliers doubling down on how they address quality matters and on-time delivery, said Warren Browne, an auto supplier consultant and former General Motors Co. executive who worked at the carmaker for 40 years. Ultimately, though, he said: "Metrics will always trump AI for whatever that means."
There may be future opportunities for growth in generative AI's application after learning from multiple purchasing rounds, said Dave Andrea, a principal and automotive strategist at accounting and management consulting firm Plante Moran. Opportunities may include forecasting risks with respect to weather disruptions and logistics or predicting foreign exchange rates.
"Because purchasing is such a complex equation with a lot of unknowns," Andrea said, "maybe through a predictive model, you can do a better analysis to reduce future costs and unintended risk."
Andrea, who conducts an annual survey on automaker-supplier relationships, said the value of partnerships ultimately is earned during the times for which AI can't be programmed. That includes sharing relevant information, accountability to commitments and fairness.
"Think about the last two to three years with pandemics, chip shortages, and all those other elements," he said. "It's just not in the playbook. That’s where you still need the relationships for the workarounds to fix what you couldn’t plan for."
GM declined to comment on whether it's using AI in its supplier selection process, citing competitive reasons. Stellantis NV isn't using AI in making supplier decisions, spokesperson Jodi Tinson said in an email.
Toyota Motor Corp. spokesperson Rick Bourgoise said in an email the Japanese automaker for years has been using AI tools to improve internal processes and is exploring applications for data assembly and analysis. The Japanese automaker, however, doesn't use and doesn't anticipate using AI for sourcing or pricing with its suppliers.
"This approach," Bourgoise said, "could have potential conflicts with Toyota’s desire to sustain long-term, mutually beneficial relationships with its supplier partners."
Mercedes-Benz Group AG has been using AI for many years in various projects along the entire value chain and has AI principles to address responsible use, privacy protection, reliability and bias, according to a statement. The Mercedes-Benz Direct Chat helps some employees do tasks more efficiently.
The automaker is also a member of the Catena-X network that connects companies across industries for secure data exchange. It's a faster process to exchange information, and the software can help detect bottlenecks at an early stage, according to the statement.
France-based supplier Forvia SE says it's using AI to respond more quickly to changes in automakers' requests for proposals. Identifying those changes alone can take hours even before responding, but AI can find the differences much faster, spokesperson Misty Matthews said.
"It just saves a lot of time," she said. "It doesn’t replace people. It gives people more time to do the revision. Sometimes, there can be a pretty tight turn."
Derek de Bono, Valeo's software defined vehicle product vice president and group product marketing vice president, says AI has some uses but not in purchasing.
France-based Valeo SE is also using artificial intelligence to make operations more efficient and faster across various departments. But like others in the industry, it doesn't see the technology completely taking over decision-making.
"AI is making it more efficient, faster, improving the process, but actually replacing a purchaser?" said Derek de Bono, Valeo's software defined vehicle product vice president and group product marketing vice president. "I don't see it on our side."
Collin Shaw, president of the MEMA trade organization for original equipment suppliers, says AI can help free up employees to spend more time relationship-building.
Many suppliers have found the most success while rolling out AI deeply in a single area first versus over multiple departments, said Collin Shaw, president of the MEMA trade organization for original equipment suppliers. Many have focused initial efforts on marketing and manufacturing, such as improving quality over purchasing. Suppliers still emphasize the importance of access to their automaker customers to build relationships, especially after the pandemic made that difficult to do.
"If it's used to parse through data and go through a lot of information, and then people's time is freed up to build relationships and work on more strategic things," Shaw said, "I think that could be a huge benefit."
